Facebook Says It’s Policing Fake Accounts, but They’re Still Easy to Spot • Hacker News Board

Re: Facebook Says It’s Policing Fake Accounts, but They’re Still Easy to Spot

#61

Post by jsnell » Mon, Nov 06, 2017, 9:21 PM UTC

cryptoz wrote: post #4

A real shocker for me is how little the big tech companies are using their AI powers for defensive purposes. Each smart-project they undertake to detect your friends in your photos and identify cats in pictures is done for a directly profitable feature. But with Twitter and Facebook both getting caught letting foreign actors buy political ads and use fake personas to influence the population, it seems like an existen…

Just how do you know what these companies are using for fighting abuse (fake accounts, spam, account hijacking, etc)? Dealing with abuse is an adversarial process, where both sides are trying to adapt to what the other is doing. It's not like most programming where you can come up with a provably good enough solution. It also means that time is a valuable resource.

There's nothing to be gained by e.g. Twitter announcing that they're now using ML on the profile images to detect fake accounts. That'd let the bad guys know right away that they need to change their pattern. Instead what you want is for them to first waste some time before they realize that something is wrong, and then waste more time experimenting on exactly what changed, and then even more time figuring out how to bypass it.

So unless you're working on abuse at one of these companies, I don't understand how you can so confidently talk about what they are and aren't doing.

Re: Facebook Says It’s Policing Fake Accounts, but They’re Still Easy to Spot

#62

Post by dymk » Mon, Nov 06, 2017, 9:22 PM UTC

ksk wrote: post #50

nerfhammer wrote: post #41

Earlier quoted context omitted.

1) ANY single prominent example of a false positive or false negative will enrage some community. People's brains can only think of examples, not rates. 2) With billions of accounts zero examples of error is an absurdity. 3) different communities drastically disagree on which things false positives and false negatives should be defined as. 4) More perceived effort to work on (1) increases the outrage, i.e. the percei…

I would think its worth trying out something even if the result is less than perfect.

So you're falling into (3) and ignoring how important (1) is

Re: Facebook Says It’s Policing Fake Accounts, but They’re Still Easy to Spot

#63

Post by twothamendment » Mon, Nov 06, 2017, 9:25 PM UTC

1024core wrote: post #52

What if this "Kevin Eversely" character had actually immigrated from Macedonia to Minneapolis? Just because many of your friends are not from your region, by itself doesn't mean anything.

Right - one odd thing alone isn't enough. I worked on a platform that allowed contributors sign up and provide content from their area. A user provided some fishy content with a very heavy slant so I started looking into his profile. The big thing that really confirmed we were on to something was his profile photo was one of a retired athlete. Busted by tineye.com. Once we dug in some more we found it was a local politician writing under a fake name.

So yes, having all your friends in another area doesn't mean you are a fake, but it may be an easy thing to look for that flags the account for more inspection.

Re: Facebook Says It’s Policing Fake Accounts, but They’re Still Easy to Spot

#64

Post by dymk » Mon, Nov 06, 2017, 9:26 PM UTC

rb2018 wrote: post #57

Facebook can just charge 1$ per sign-up and the problem is gone (faking credit cards etc does not scale). Of course, as others have said, they prefer having more "users".

Are you honestly suggesting that requiring having a credit card is in any way a solution to fake accounts? That's a ridiculous proposition not just for Facebook, but any service that operates in developing areas or serves users under the age of 18.

Re: Facebook Says It’s Policing Fake Accounts, but They’re Still Easy to Spot

#65

Post by mikestew » Mon, Nov 06, 2017, 9:41 PM UTC

twothamendment wrote: post #59

mikestew wrote: post #39

Earlier quoted context omitted.

Things mass produced cheaply are lesser quality, but we're willing to make the trade off as a society because we get things for cheaper. Problem is, from what I'm seeing, is that the market bifurcates into "shitty, but cheap" and "OMG, the price, but quality". I want the middle way: "reasonably priced, 'good enough' quality, and won't break the bank". Tools are a prime example. You had Snap-On/Matco/etc., Sears Craft…

Wandering far from Facebook (the farther the better IMHO) - Craftsman has been dead to me for years so I tried a Dewalt socket/wrench set. The name should be around a while, they have a good warranty and they feel like I shouldn't need it. It was god to see someone was trying to fill in that middle of the road "good enough" and affordable.

I’ll take a look next time I see Dewalt, thanks. Harbor Freight is almost good enough, but not quite there, so I’m eager to try some Dewalt.

Re: Facebook Says It’s Policing Fake Accounts, but They’re Still Easy to Spot

#66

Post by fredgrott » Mon, Nov 06, 2017, 10:12 PM UTC

question Can any Social Platform claim to police bot accounts when receiving investment money from the same exact forces doing the bot accounts? Seems to me that their is a bigger problem than just the fake accounts..no?

Re: Facebook Says It’s Policing Fake Accounts, but They’re Still Easy to Spot

#67

Post by ProAm » Mon, Nov 06, 2017, 10:13 PM UTC

lwis wrote: post #53

Yet I've contacted them twice about a fake profile which they've refused to remove despite passport and driving license evidence. Does anyone know of a better channel through which to pursue this?

Honestly who cares if there are fake accounts (as a user). If you are an investor I could possibly see interest, but even then who really cares as long as facebook looks like its big and still growing that's all that matters.

Re: Facebook Says It’s Policing Fake Accounts, but They’re Still Easy to Spot

#68

Post by deadmetheny » Mon, Nov 06, 2017, 10:20 PM UTC

jokoon wrote: post #19

By "policing fake accounts" they also mean "hunt down real accounts who don't use their real name, and ask for a scan of your ID".

If you don't want to abide by their policies, don't sign up. It's not that hard.

Re: Facebook Says It’s Policing Fake Accounts, but They’re Still Easy to Spot

#69

Post by erjjones » Mon, Nov 06, 2017, 11:27 PM UTC

Crowd source this problem. Add a report fake account option. If so many (x) number of different IP address, etc. report it, sandbox that account. Simple.

Re: Facebook Says It’s Policing Fake Accounts, but They’re Still Easy to Spot

#70

Post by philipwhiuk » Mon, Nov 06, 2017, 11:37 PM UTC

wpietri wrote: post #35

jmcqk6 wrote: post #28

Earlier quoted context omitted.

>First, there are the countless instances where something is trivial for a single person to suss out that doesn't scale. This is an interesting argument to me. It's basically conceding a shittier solution, because we "can't scale" the effective one. As we talk more and more about automation, I think this is similar to what we've seen in meat space with things like tools. I've heard so many people complain about chine…

Yes. And the way it's often argued, it treats the status quo as some sort of natural given, like the weather. The line of argument seems to be that if Facebook and Twitter can't run safe platforms while maintaining their current revenue models and profit margins, well, we'll just have to live with that lack of safety. I find this line of argument odd when it's about actual natural givens, like dangerous weather. I'm…

Yes, I've always found the 'YouTube gets too much video to moderate it' a particularly shit argument. Maybe they not be allowed to have as much content then.